Two possible entities of endometriosis-associated ovarian cancer: correlated or incidental?

This study aimed to describe 2 types of endometriosis-associated ovarian cancer: those with transitional elements (atypical endometriosis and borderline tumors) termed endometriosis-correlated or incidental benign endometriosis vs ovarian cancer cases not associated with endometriosis. This was a prospective, observational, monocentric study conducted from November 2021 to December 2023. Patients with ovarian cancer eligible for surgery were enrolled and classified into endometriosis-correlated ovarian carcinoma, endometriosis-incidental ovarian carcinoma, or ovarian carcinoma without endometriosis groups based on the presence or not of endometriosis and transitional lesions. Clinical, sonographic, surgical and pathological data and progression-free survival were recorded. Logistic regression models for accurate patient classification were developed from pre-surgical variables. Of the 170 patients included, 83 (48.82%) had ovarian carcinoma without endometriosis, 39 (22.94%) had endometriosis-incidental ovarian carcinoma, and 48 (28.24%) had endometriosis-correlated ovarian carcinoma. Patients with endometriosis-incidental ovarian carcinoma and endometriosis-correlated ovarian carcinoma were diagnosed at younger ages (p = .002) and had lower post-menopausal rates than patients with ovarian carcinoma without endometriosis (p = .011). Patients with endometriosis-correlated ovarian carcinoma had fewer pregnancies (p < .001) and higher CA-19.9 levels (p = .002) presented with unilateral and multilocular solid lesions than patients with ovarian carcinoma without endometriosis (p < .001). Patients with endometriosis-incidental ovarian carcinoma showed intermediate lesion morphology. Endometriosis-correlated ovarian carcinoma was mostly diagnosed at early Federation of Gynecology and Obstetrics stages (range; I-II) compared with endometriosis-incidental ovarian carcinoma and ovarian carcinoma without endometriosis (p < .001), had less extensive disease (p < .001), and a higher likelihood of complete cytoreduction (p = .035). Endometriosis-correlated ovarian carcinoma was more likely to include clear cell, endometrioid, and mesonephric-like adenocarcinomas, whereas serous histotype predominated in the other groups (p < .001). Logistic regression models accurately identified patients with endometriosis-correlated ovarian carcinoma vs patients with endometriosis-incidental ovarian carcinoma (area under the curve [AUC] = 0.926) and ovarian carcinoma without endometriosis (AUC = 0.968) but could not reliably differentiate endometriosis-incidental ovarian carcinoma from ovarian carcinoma without endometriosis (AUC = 0.668). The 2-year progression-free survival rates were 91% in endometriosis-incidental ovarian carcinoma, 80% in endometriosis-correlated ovarian carcinoma, and 59% in ovarian carcinoma without endometriosis (p = .024). Our study indicates that ovarian cancer associated with endometriosis consists of 2 clinical entities, with endometriosis-incidental ovarian carcinoma emerging as a bridging group between endometriosis-correlated ovarian carcinoma and ovarian carcinoma without endometriosis.